| v0.1 |
Past: Initial beta release in R. Established the SVEIWRS compartmental structure, the mobility and environmental transmission terms, and the observation process. Used to scope data sources and parameter priors prior to the LASER port. |
| v1.0 |
Current: First production implementation. The transmission engine was developed in the Python laser-cholera metapopulation model and has since been ported to pure R (run_simulation()) inside the R MOSAIC package, which also handles data assembly, parameter sampling, and Bayesian calibration on a local parallel cluster. Priors for all biological parameters are now literature-derived (shedding, ID50, decay, CFR), and the calibration pipeline produces medoid and ensemble forecasts with importance-sampling weights, effective sample size, and agreement-index diagnostics. |
| v2.0 |
Future: Maximizing the metapopulation approach. District-level data and improvements to model fitting via hierarchical likelihoods. |
| v3.0 |
Future: Agent-based component with richer immune dynamics, including biphasic vaccine waning and individual-level exposure history. |